Impact of prosodic structure and information density on vowel space size

Abstract

We investigated the influence of prosodic structure and information density on vowel space size. Vowels were measured in five languages from the BonnTempo corpus, French, German, Finnish, Czech, and Polish, each with three female and three male speakers. Speakers read the text at normal, slow, and fast speech rate. The Euclidean distance between vowel space midpoint and formant values for each speaker was used as a measure for vowel distinctiveness. The prosodic model consisted of prominence and boundary. Information density was calculated for each language using the surprisal of the biphone Xn|Xn−1. On average, there is a positive relationship between vowel space expansion and information density. Detailed analysis revealed that this relationship did not hold for Finnish, and was only weak for Polish. When vowel distinctiveness was modeled as a function of prosodic factors and information density in linear mixed effects models (LMM), only prosodic factors were significant in explaining the variance in vowel space expansion. All prosodic factors, except word boundary, showed significant positive results in LMM. Vowels were more distinct in stressed syllables, before a prosodic boundary and at normal and slow speech rate compared to fast speech.

Dimensions of Segmental Variability: Interaction of Prosody and Surprisal in Six Languages

“…In contrast to Schulz et al (2016) and Aylett and Turk (2006), we found a positive but non-significant effect of stress on vowel dispersion. This difference might be due to the weak positive correlation between surprisal and stress (r = 0.23).…”

“…Effects for both variables cannot be fully separated in a statistical model. In addition, Schulz et al (2016) analyzed only five languages of the BonnTempo corpus, DEU, CES, POL, FIN, and FRA. The present study also includes (American) English, a language which shows a relatively strong correlation between surprisal and vowel dispersion [r = 0.26, t (558) = 6.39, p < 0.001].…”

The diachronic development of combining forms in scientific writing

“…Texts of more recent scientific articles typically contain a higher average information density than texts from earlier periods (Degaetano-Ortlieb et al an influence on the expectations of text users with regard to upcoming linguistic elements based on previous experience with similar texts. The effect of information density on choices between different types of phonological or syntactic alternatives (e.g., the preferences of full or contracted auxiliaries or the use of complement clauses with and without the complementizer 'that') has been addressed by various scholars, especially in psycholinguistic studies (Demberg & Keller 2008;Hale 2001;Levy 2008) and recently also in corpus-linguistic studies (e.g., Degaetano-Ortlieb & Teich 2017; Schulz et al 2016;Zimmerer et al 2017). However, morphological structures within complex lexemes with the potential to combine various lexical morphemes still remain to be studied.…”

Mel-Cepstral Distortion of German Vowels in Different Information Density Contexts

“…Conventions and rules of prominence, as well as predictability of linguistic units are language-specific. Also, the influence of information density on phonetic encoding density seems to be language-specific [4].…”

“…Phonetic structures in high ID contexts are increased, while reduced phonetic structures are usually found in low ID contexts. Easily predictable vowels are less dispersed than vowels that are less predictable [3,4]. Vowel dispersion is defined as the Euclidean distance between the average centre of the vowel space and formant values for each vowel [5].…”